{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"submitter":["Zhai J"],"funding":["NIA NIH HHS","National Institutes of Health","NIGMS NIH HHS","NIH HHS","National Science Foundation"],"pagination":["bbaf212"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC12078937"],"repository":["biostudies-literature"],"omics_type":["Unknown"],"volume":["26(3)"],"pubmed_abstract":["Catalytic constant (Kcat) is to describe the efficiency of catalyzing reactions. The Kcat value of an enzyme-substrate pair indicates the rate an enzyme converts saturated substrates into product during the catalytic process. However, it is challenging to construct robust prediction models for this important property. Most of the existing models, including the one recently published by Nature Catalysis (Li et al.), are suffering from the overfitting issue. In this study, we proposed a novel protocol to construct Kcat prediction models, introducing an intermedia step to separately develop substrate and protein processors. The substrate processor leverages analyzing Simplified Molecular Input Line Entry System (SMILES) strings using a graph neural network model, attentive FP, while the prote"],"journal":["Briefings in bioinformatics"],"pubmed_title":["NNKcat: deep neural network to predict catalytic constants (Kcat) by integrating protein sequence and substrate structure with enhanced data imbalance handling."],"pmcid":["PMC12078937"],"funding_grant_id":["R01 GM149705","R01GM149705","1955260","R01 AG057555","R01AG057555"],"pubmed_authors":["Liu Y","Qi X","Xie L","Cai L","Zhai J","Wang J","Tang H"],"additional_accession":[]},"is_claimable":false,"name":"NNKcat: deep neural network to predict catalytic constants (Kcat) by integrating protein sequence and substrate structure with enhanced data imbalance handling.","description":"Catalytic constant (Kcat) is to describe the efficiency of catalyzing reactions. The Kcat value of an enzyme-substrate pair indicates the rate an enzyme converts saturated substrates into product during the catalytic process. However, it is challenging to construct robust prediction models for this important property. Most of the existing models, including the one recently published by Nature Catalysis (Li et al.), are suffering from the overfitting issue. In this study, we proposed a novel protocol to construct Kcat prediction models, introducing an intermedia step to separately develop substrate and protein processors. The substrate processor leverages analyzing Simplified Molecular Input Line Entry System (SMILES) strings using a graph neural network model, attentive FP, while the prote","dates":{"release":"2025-01-01T00:00:00Z","publication":"2025 May","modification":"2026-07-01T03:21:42.339Z","creation":"2025-09-01T03:04:57.832Z"},"accession":"S-EPMC12078937","cross_references":{"pubmed":["40370097"],"doi":["10.1093/bib/bbaf212"]}}